Comparison of Models Fitting Alternative Contemporary Group Definition and Data Editing Procedures in the Genetic Evaluations of Reproductive Performance and Cow Longevity Traits in Nellore Cattle

Reproductive traits in beef cattle are strongly influenced by environmental and management factors, making the correct definition and quality control of contemporary groups (CG) a key step for reliable genetic evaluations. Therefore, the main objective of this study was to evaluate the impact of alternative models considering different definitions of CG, minimum CG sizes (MCGS), and standard deviation thresholds for phenotypic outlier removal on the genetic evaluation of cow reproductive and longevity traits in Nellore cattle (Bos taurus indicus). The four traits evaluated included age at first calving (AFC; in days), early pregnancy occurrence (EP; binary), scrotal circumference at 18 months (SC18; in cm), and total number of calvings at 53 months of age (NC53; count). The database comprised 954,216 animals, including 401,885 females and 552,331 males, with 227,447, 401,885, 356,519, and 531,579 records available for AFC, EP, NC53, and SC18, respectively, and a pedigree file comprising 1,348,164 animals. Variance components and genetic parameters were estimated using a pedigree-based animal model implemented in the BLUPF90 software family (version 1.57) for continuous traits (AFC and SC18) and a Bayesian animal model implemented in THRGIBBS1F90 (version 2.108) for categorical traits (EP and NC53). The Linear Regression (LR) method was used to assess model performance based on accuracy, bias, and dispersion of estimated breeding values (EBVs). The main CG criteria included MCGS of 3, 7, and 15 animals, four phenotypic outlier-removal thresholds (±2.0, ±2.5, ±3.0, and ±3.5 SD) for continuous traits, and the presence or absence of phenotypic variability within CGs for categorical traits. We also compared alternative combinations of fixed effects included within or outside CGs and evaluated genetic connectedness among CGs. The definition of CGs and data editing procedures yielded different heritability estimates and EBV accuracies across scenarios and traits. The heritability estimates for the evaluated traits ranged from 0.06 to 0.45, while EBV accuracies ranged from 0.45 to 0.92 across CG scenarios. Continuously-recorded traits, especially AFC, were more sensitive to model specification and CG structure, whereas categorical traits and SC18 were more robust to differences in quality control criteria. Animal rankings showed high consistency among scenarios, with mean concordance of 94.8% for the top 1% of males and top 20% of females and 95.4% for all animals. Our findings indicate that the definition of CGs and data editing procedures for each trait and population are key steps to be carefully performed before genetic evaluation analyses, since parameters such as heritability estimates, EBV accuracy, bias, and dispersion may be affected.